# Multi-image Medical Visual Question Answering on MIM-RAD (Held-out) Source: [Multi-image Medical Visual Question Answering on MIM-RAD (Held-out)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:39:24.631Z Evaluation of multi-image visual question answering performance on the held-out split of the MIM-RAD dataset, reporting Close (C) and Open (O) scores. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Multi-image Medical Visual Question Answering | | Dataset (including any stated split) | MIM-RAD (Held-out) | | Metric | Close Score (C) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 10 of 10 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Close Score (C)](): Higher is better; unit: PERCENT. - [Open Score (O)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Close Score (C) | Result date | Report paper | | --- | --- | --- | --- | | GPT-4o | 0.69 | Not recorded | 68fa78c01e209131cea7a40c | | Med-Mantis | 0.6533 | Not recorded | 68fa78c01e209131cea7a40c | | Mantis-8B | 0.5967 | Not recorded | 68fa78c01e209131cea7a40c | | InternVL2-8B | 0.5633 | Not recorded | 68fa78c01e209131cea7a40c | | MIM-LLAVA-Med | 0.5333 | Not recorded | 68fa78c01e209131cea7a40c | | LLaVA-Med-7B | 0.4867 | Not recorded | 68fa78c01e209131cea7a40c | | deepseek-VL-7B | 0.17 | Not recorded | 68fa78c01e209131cea7a40c | | Mantis† | 0.0566 | Not recorded | 68fa78c01e209131cea7a40c | | LLaVA-Med† | 0.0466 | Not recorded | 68fa78c01e209131cea7a40c | | Med-Flamingo-9B | 0.04 | Not recorded | 68fa78c01e209131cea7a40c | ### GPT-4o Close Score (C): 0.69. Result ID: 69959fae3badfaffe69c1961. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Med-Mantis Close Score (C): 0.6533. Result ID: 69959fae3badfaffe69c1969. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"training\_dataset":"Med-MIM instruction dataset"}. This record does not assert independent reproduction. ### Mantis-8B Close Score (C): 0.5967. Result ID: 69959fae3badfaffe69c1966. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### InternVL2-8B Close Score (C): 0.5633. Result ID: 69959fae3badfaffe69c1964. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### MIM-LLAVA-Med Close Score (C): 0.5333. Result ID: 69959fae3badfaffe69c1967. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"training\_dataset":"Med-MIM instruction dataset"}. This record does not assert independent reproduction. ### LLaVA-Med-7B Close Score (C): 0.4867. Result ID: 69959fae3badfaffe69c1965. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### deepseek-VL-7B Close Score (C): 0.17. Result ID: 69959fae3badfaffe69c1963. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Mantis† Close Score (C): 0.0566. Result ID: 69959fae3badfaffe69c196a. [Reporting paper]() Recorded evaluation configuration: {"setup":"without composed instruction dataset"}. This record does not assert independent reproduction. ### LLaVA-Med† Close Score (C): 0.0466. Result ID: 69959fae3badfaffe69c1968. [Reporting paper]() Recorded evaluation configuration: {"setup":"without composed instruction dataset"}. This record does not assert independent reproduction. ### Med-Flamingo-9B Close Score (C): 0.04. Result ID: 69959fae3badfaffe69c1962. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.